[Derek in NYC]

A few thoughts on your criteria.

First, while I have done so in a B&M club, I have never made a table selection decision online based on relative position. Fact is, it is hard enough finding a decent table with more fish than sharks, I can't be too picky about my relative seat position if I can get on a good table. Additionally, whether you sit to the right or left of a TAG/LAG/LP whatever isn't going to determine whether you are going to win or lose... you just need to use different skills such as reraising to isolate, checkraising the field for value, etc. I personally think that learning how to use these skills is key to improving your game. If one's profitability turns on having ideal relative position, I propose that such a player's game is a crippled, mechanical game that won't be able to grow into higher limits.

One factor that I personally consider important, but that wasn't mentioned in your list, is stack size. Setting aside the sharks/fish ratio, I consider it highly desirable when all players have at least 15-20 BBs in front of them. (Obviously this is better when the fish have huge stacks, but it is generally applicable.) Deeper stacks means more stable opponents, which means better reads over the course of a session. Deeper stacks also means more stable tables that dont break down into shorthanded sessions. I will generally avoid sitting if more than 3 players have microstacks.

Absent buttons. Since Im sitting down at a full ring, obviously the presence of absent buttons is a negative. The blinds circulate faster, and I'm forced to play a game where I have less of an edge. So I won't sit if there are 2 or more absent buttons. Another reason I will avoid these tables is that in my experience they tend to be unstable. At around 3-4 absents, a table completely breaks down and people stop autoposting blinds. Next thing you know, you're heads up. In fact, as a table starts to empty out or go absent, I will unclick autopost and make each blind posting decision individually.

As you can see, a lot of my factors involve table stability. I hinted at the reason for this, but I'll be explicit. Stable tables give you good reads. You have a solid pokertracker dataset for everyone at the table, and you can maximize your edge as a skilled player. Where the table is unstable, and people are popping in and out (short stacks, absent buttons, etc.), you are forced to play a much more mechanical game. Harder to defend blinds, harder to 3-bet for isolation, harder to induce bluffs, etc. Overall, I think table stability is very important to a profitable session.